The present invention relates to a sound recording system particularly adapted to create a recording which can produce a highly dimensionalized impression of the sound.
To explain how prior art high quality stereophonic recordings are made, reference is made to FIG. 1 where there are shown two microphones M1 and M2 spaced from each other and positioned in front of an orchestra, indicated generally at 10, and comprising a plurality of orchestral components (i.e. musical instruments 12a through 12g). Let it be assumed that the two microphones M1 and M2 are positioned relatively close to the orchestra and spaced from each other by moderately more than ten feet.
The sound from instrument 12f travels on a first shorter path 14 to microphone M2, and on a second longer path 16 to microphone M1. Obviously the microphone M2 will record the sound from instrument 12f at a higher intensity than will the microphone M1. Also, there will be a phase shift in that the microphone M1 may pick up the sound in the order of approximately 1/100 of a second later than the microphone M2, since the sound must travel further to the microphone M1.
In like manner, the sound emanating from the instrument 12b would reach the microphone M1 sooner and at a higher intensity than the sound from the instrument 12b would reach the microphone M2. At some intermediate location, e.g. at the location of the instrument 12d, the sound would reach the two microphones M1 and M2 at approximately the same time and same intensity. The sound transmitted to the microphone M1 is used to produce a first signal which is transmitted to the "STEREO RECORDER" and in turn is reproduced in the sound recording (i.e. a tape or a phonograph record) in a manner that when the recording is played, this signal, corresponding to the sound at microphone M1, reproduces in one speaker a sound which is a substantial reproduction of the sound reaching the microphone M1. In like manner, the sound reaching the microphone M2 is used to produce a signal which is also recorded on the sound recording in a manner that when the recording is played, the signal driving the other speaker corresponds to the sound transmitted to the microphone M2.
When the stereophonic recording is played in a typical stereophonic sound reproduction system, there are two speakers positioned at two laterally spaced locations. The listener is positioned rearwardly of the speakers and facing toward a location between the speakers. A distinct sound that is transmitted only from the left speaker can be detected by the listener as coming from that source since the left and right ears of the listener will detect a difference in intensity and also detect a phase shift so as to obtain the impression of the direction of the sound. When another distinct sound is transmitted from the right speaker, the direction of that source of sound can also be detected by the listener. Thus, the sound can be expanded to the area encompassed by the two speakers.
With reference being made to FIG. 2, let it be assumed that the signal produced at the microphone M2 is used in a manner to produce sound in the left speaker LS, while the signal produced at the microphone M1 is used in the recording to produce a sound in the right speaker RS.
To relate this specifically to the recording made in accordance with the arrangement of FIG. 1, with regard to the sound produced by the instrument 12f, the speaker LS will first reproduce this sound at the higher intensity level. This sound would have first and second components Ll and Lr reaching the left and right ears, Le and Re, respectively. At that instant, the two ears Le and Re would detect a difference in intensity of the two sound components Ll and Lr and also a phase shift of possibly 100 to 300 microseconds so that there would be a very definite sense of direction from the left speaker LS. However, about 1/100th of a second later, essentially the same sound produced originally from the instrument 12f would be reproduced from the right speaker RS at a lower level of intensity, along the two path components Rl and Rr. If the sounds from the two speakers LS and RS were of equal intensity, there would be no stereophonic effect. However, with the sound from the left speaker LS being of greater intensity, there is something of the stereophonic effect, but this is obscured to some extent by a very similar sound emanating from the right speaker RS.
Consideration is now given to the sounds emanating from the centrally located instrument 12d. As indicated previously, since the distances from the instrument 12d to the two microphones M1 and M2 are substantially equal, the timing and intensity of the sounds at M1 and M2 are substantially the same.
Thus, when the sound from the instrument 12d is reproduced in the two speakers LS and RS, the sounds from the two speakers LS and RS traveling the main path components Ll and Rr reach the left and right ears Le and Re simultaneously. The two secondary sound components, Lr and Rl, reach the two ears Le and Re simultaneously, but possibly 100 to 300 microseconds later than the main sound components Ll and Rr. The overall effect is that this sound component appears to emanate from a location between the speakers. Thus there is the overall stereophonic effect of sound coming from the speakers and from areas between the speakers.
There have been attempts in the prior art to give even greater dimension to the sound reproduction system, so that there is the impression that the sound is coming from areas totally outside of the more limited area at and between the two speaker locations. While the applicant is not totally familiar with the operation of these systems, according to the applicant's present understanding, such systems require rather limited conditions of operation. For example, it is known that the prior art systems known to the applicant must be utilized in an environment where there is very little reflected sound, for example in an open space, or in a room where the walls are made of a highly sound absorbent material. Further, the systems which are known to the applicant are quite sensitive to the location of the hearer's head. Thus, if the person moves his head from a precise listening location, or rotates his head moderately toward one speaker or the other, a large part of the dimensionalized effect is lost. Thus, to the best knowledge of the applicant, these systems have remained more in the category of laboratory curiosities, rather than a system which is practical for general use.
It is believed that the prior art systems discussed immediately above are operated on the basis of recognizing that sound emanating from various locations both forwardly and rearwardly of a person's head create different sound patterns relative to the person's ears. Thus, a sound emanating from a location in front of the person and 30.degree. to the left would produce distinctly different relative sound patterns to the person's ears than a sound emanating directly from a location at the person's left. There would be a difference in intensity for the various frequencies, and also a different phase shift detected by the person's two ears. It is believed that this phenomenon is utilized to tailor or control the sound emanating from the speakers to cause delicate adjustments in the phase shift and sound intensity at different frequencies to produce the effect of greater dimensionalized sound. However, as indicated above, it is believed that the sensitivity of such systems to reflected sound and also head location have not made them practical for general use.
Also, there has been in the prior art recognition of the phenomenon called "cross talk" which in certain circumstances has the effect of degrading the quality of the sound transmitted from two spaced speakers. To describe this phenomenon briefly, the sound from a right speaker reaches both the right and left ear of the person, but reaches the left ear at a slightly later time depending on the distance between the speakers, the listening angle, and the ear spacing of the listener (e.g. at a time ranging from zero to 900 microseconds) and at a somewhat lower intensity than the sound which reaches the right ear. The sound from the left speaker acts in somewhat the same way relative to the left and right ears. With similar sounds being emitted from both speakers something of the stereophonic effect is lost or at least diminished by this phenomenon of cross talk.
This problem was recognized in U.S. Pat. No. 4,058,675, and this patent disclosed a system which has for its intended purpose the elimination of the deteriorating effect of cross talk. Since it is believed that a deeper understanding of the apparatus of U.S. Pat. No. 4,058,675 will aid in a fuller appreciation of the present invention, the apparatus will be discussed in some detail herein.
Reference is again made to FIG. 2, which shows the left speaker LS and the right speaker RS at two spaced locations, and the person P at a listening location equally distant from the speakers LS and LR and located rearwardly of the speakers. The person has a left ear Le and a right ear Re.
The sound from the left speaker can be considered as having two components, namely component Ll which is transmitted from the left speaker LS to the left ear Le, and a second component Lr which is transmitted from the left speaker LS to the right ear Re. The right speaker in like manner has two sound components Rr and Rl transmitted to the right ear Re and left ear Le, respectively.
In U.S. Pat. No. 4,058,675, there is a discussion of the effect of cross talk in that the sound reaching the right ear Re from the right sound component Rr reaches the right ear shortly before the left sound component Lr reaches the right ear Re. Thus, if substantially the same sound is being transmitted from the left and right speakers, the right ear will hear the sound first at a higher intensity, and the same sound with a slightly delayed phase shift at a lower intensity. U.S. Pat. No. 4,058,675 proposes to alleviate this problem by providing additional left and right auxilliary speakers to provide cancelling sounds to eliminate cross talk. This will be explained with reference to FIG. 3, labelled "Prior Art" and corresponding to FIG. 5 of U.S. Pat. No. 4,058,675.
It can be seen that there is a left speaker LSP, made up of a main left speaker MSL and a left subspeaker SSL. The left signal L enters at the terminal "IN", and is transmitted directly to the main left speaker MSL. In addition, the left signal is applied through an attenuator AT and phase shift and delay means DP to the subspeaker SSL. In like manner, the right speaker SPR has a main speaker MSR and subspeaker SSR, along with an attenuator AT and delay means DP.
In the operation of the device of U.S. Pat. No. 4,058,675, the main sound component Ll from the main left speaker MSL reaches the left ear Le substantially undiminished. Also, there is a main sound component (not shown herein for clarity of illustration) from the main right speaker MSR to the right ear Re. However, the second sound component Rl from the right main speaker reaches the left ear El slightly later than the main right sound component reaches the left ear and also later than the left main sound component Ll reaches the left ear. The delayed and inverted signal SLl from the left subspeaker SSL is timed at a predetermined phase shift and directed at a predetermined intensity to substantially cancel the right second component Rl. Thus, the cross talk from the right speaker MSR is substantially attenuated. The operation is substantially the same with respect to the right ear, so that the left ear hears sounds mainly from only the left speaker, while the right ear hears sounds mainly from the right speaker.
While the applicant has not conducted an exhaustive analysis of the device shown and described in U.S. Pat. No. 4,058,675, the analysis and limited experimental evaluation which was done indicates that such apparatus has significant limitations in producing any dimensionalized sound effect beyond that which is obtained from the conventional stereo system. To explain this more fully, let it be assumed that the prior art system shown in FIG. 3 is used to play the recording made from the system shown in FIG. 1.
Let it now be assumed that with respect to the reproduced sound corresponding to that produced from the instrument 12d, the two secondary sound components Lr and Rl are both cancelled, so that the Le hears only a left main sound component Ll and the right ear Rr hears only the right main component Rr. With the two ears hearing substantially the same sound at the same intensity, there will be the impression that the sound is coming from a central location immediately forward of the person P. In the apparatus of the U.S. Pat. No. 4,058,675, it is possible to obtain the cancellation of the secondary sound components Lr and Rl, where the sound corresponds to the sound emanated from the instrument 12d (i.e. where the sound from the two speakers LS and RS are in the same phase relationship and at the same intensity). However, where the two sounds correspond to those originating from the instrument 12f and 12b, (where the two sound components are reproduced at substantially different intensity and with one sound being delayed substantially from the other), there would be no cancellation of the secondary sound components Lr and Rl by use of the apparatus of U.S. Pat. No. 4,058,675 since the inverted and delyed sound from the sub speaker would be so far out of phase from a corresponding sound from the opposite main speaker. However, since there is a substantial difference in sound intensity, there would still be something of the stereophonic effect when using the speaker system of U.S. Pat. No. 4,058,675, in the same manner as with a typical stereophonic system.
In view of the foregoing, it is an object of the present invention to provide a highly dimensionalized. sound recording, and an apparatus and method of making the same.